Kaposi's sarcoma-associated herpesvirus-8 (KSHV) plays a clear causative role in the pathogenesis of Kaposi's sarcoma (KS). KS is most frequently found in immunosuppressed patients such as patients with acquired immunodeficiency syndrome. KSHV elicits stronger HLA CD8+ cytotoxic T-lymphocyte (CTL) responses in healthy adults than in HIV-infected individuals, supporting the potential of vaccines as a strategy for controlling KSHV-related diseases. We propose an approach toward the development of a KSHV vaccine based on the stimulation of CD8+ CTL responses more efficiently than the natural infection. This will be accomplished by using synthetic combinatorial libraries (SCLs) to identify peptides that would be effective as immunogens in stimulating T-cell mediated immune responses against KSHV infection. The SCL approach, particularly when generated in a positional scanning (PS) format, allows the direct identification of the key residue(s) of active peptide sequence(s) from the library screening. Each mixture making up nonapeptide and decapeptide PS-SCLs will be screened for their ability to stimulate cytokine production and/or cytolytic activity by CTL clones derived from HLA-A2-positive patients that recognize targets in kaposin (ORF K12), LANA (ORF 73), and SCIP (ORF 65) proteins and have lytic specificity for KSHV-infected cells. The screening results will also be systematically compared to nonapeptides or decapeptides in viral or human proteins in databases to predict and identify stimulatory natural peptides using a biometrical analysis. Based on the screening results and biometrical analysis, series of individual peptides will be synthesized and their ability to stimulate T-cell mediated immune response determined. The peptides immunogenicity will be assessed by performing in vivo immunization using the HLA-A2.1 Kb mouse model, and further evaluated through in vitro stimulation experiments. The cross-recognition of the most potent peptides will then be investigated using a cohort of HLA-A2-positive patients and their cross-reactivity determined by challenging T-cells generated through the immunization or stimulation studies to the native ligand. The PS-SCL approach combined with the biometrical analysis will allow 1) the generation of optimal epitope mimics, 2) the identification of natural ligands in those proteins for which the optimal epitope is unknown, and 3) provide insight into potential cross-reactivity between different viral or human protein sequences and KSHV epitopes.